Demystifying RPA Myths and Misconceptions
- By Winston Thomas
- July 22, 2020
Mention robotic process automation (RPA) and many see software bots making human workers obsolete.
Nothing can be further from the truth. In fact, RPA has helped to create jobs and make human workers more productive.
More often a company is deploying RPA to help their human workforce to handle peaks or rising workloads without having to increase their hires. In other cases, RPA projects are there to help human workers ascend the value chain to increase their contribution to the company and not be weighed down by the manual, repetitive tasks.
These were some observations that were shared and discussed at a recent CDOTrends Digital Dialogue session, organized with Automation Anywhere. The discussion, held under Chatham House Rule, saw Siow E Hunt, ASEAN partner lead at Automation Anywhere and Madhu Mani, principal consultant at AsiaPac, an M1 company lead the discussion. They were joined by:
- Gautham Bartake, senior IT manager for services integration in Asia at Bayer
- Lee Giok Leng, head of IT at GuocoLand Ltd
- Domnick Almeida, senior director for IT planning and architecture in Asia Pacific at DHL
- Rajeswar Mothe, chief technology officer at Roboxa
- Arnold Leong, executive manager at Building and Construction Authority
- Sandra Lam, vice president for digital-led partnerships at Citibank
- Elvina Melissa, application analyst at Daimler
- Govind Khandelwal, director of cyber strategy and planning at Standard Chartered
The lively discussion delved into popular misconceptions about RPA and asked key questions. The answers are as below.
Can RPA automate everything?
The best way to understand what RPA is good at is to know what processes will benefit the most. Rule-based processes — those that are highly repetitive, manual and only require minimal human decisions — are the biggest beneficiaries.
Error-prone manual processes are also good candidates. RPA can automate the processes where there is a lot of manual copying and pasting (like transferring values between spreadsheets or merging them), and which can introduce human errors. This is the reason RPA projects often begin within administration, finance and procurement departments.
Type of data is crucial. RPA is designed for working with structured data. For semi-structured data, such as files and PDF documents, Automation Anywhere has a bot to capture the data in structured data format.
Ensuring the right return on investment is equally important. Starting with processes where automation can offer the most savings in terms of time, effort and cost will be best and offers a strong case to deploying RPA in other parts of the organization.
One area where companies overlook is using RPA for processes that are part of a system that will eventually be replaced or made redundant. Then the bots will need to be designed and trained when the new system is deployed. Better to begin after a system upgrade or replacement.
How do you address user resistance?
One of the biggest challenges for RPA is human resistance. Mention automation and participants agreed that employees think their jobs are at risk.
The problem comes down to transparency. Participants shared that when the top management is not clear about the intent of the RPA project, employees get defensive.
Clarity of what RPA can do is another issue. People mistakenly think RPAs can take over a job 100%. It cannot as there will always be aspects where human decisions are needed. Bots can however reduce the workload and allow their human counterparts to focus on more complex decisions. Nevertheless, this misconception can lead to disappointment and create resistance to new RPA projects.
Participants agreed that these misconceptions need to be addressed first by making communication and intent clear right from the onset. And instead of getting management to highlight the benefits of the project, they should work with ambassadors or champions within the user communities.
One participant also noted that progress and job loss should not be a zero-sum game. Project proponents should also highlight how RPA can help users improve their productivity in their own terms. For example, one can highlight how much time an accountant can save from doing routine, mundane tasks.
Should IT or business lead the RPA project?
While RPA targets business processes, it needs technical expertise to understand how it can be made to work in the current business environment. After all, no two environments are exactly the same. The advice is to have both IT and business work together on RPA projects. This way, each can identify areas that the other misses.
The answer also depends on how an organization is structured and how IT works with the business departments. Participants shared that there are cases where proactive IT teams deployed RPA projects to help their business teams. Conservative IT teams have had the opposite effect.
Instead of waiting for the right team to lead, it may be better to start with the outcome. If the outcome — for example, business productive gains by a certain percentage — is clear, then both IT and business can work together to achieve it.
Can you audit RPA?
Automation Anywhere makes it possible to audit the bot’s actions. It is a key strength of the company’s value proposition that includes a control room for full visibility of the bot actions.
This is vital for regulated industries like financial services and utilities where regulators want to know whether the bot behavior is following established protocols and are making compliant decisions.
The same features can also help companies to reinforce security when scaling the number of bots to hundreds. It can highlight if a single bot (out of a wide range and number) is acting erratically or being compromised.
What are some popular use cases?
The use cases for RPA are diverse. The popular ones are in finance, administration and procurement/logistics. The primary reason is that these functions include numerous repetitive and manual tasks.
One example for administration is creating sales orders. Let’s say a company had to create 400 sales orders at the end of each month because of a recurring process. Creating these can take a few days and is error prone as it can involve a lot of copying and pasting. As a result, one company had to frequently raise credit notes. RPA automated the entire process, eliminated the need to raise credit notes, and the staff only had to handle the few exceptions.
Another example is Finance streamlining accounts receivables (AR). Generating an AR summary for each customer because of overdue payments can be laborious. RPA can automate the entire process right down to sending the emails.
RPA starred well during COVID-19. One organization used bots to scan and collate information. In another case, a bank used bots to quickly roll out loan deferment processes for their users.
In the past, these projects would have required an army of consultants and a lot of process re-engineering. RPA only requires two to three programmers.
Winston Thomas
Winston Thomas is the editor-in-chief of CDOTrends. He likes to piece together the weird and wondering tech puzzle for readers and identify groundbreaking business models led by tech while waiting for the singularity.